PRIMAL stands for a Pipeline of Root Image analysis using MAchine Learning. In short, the pipeline use Machine Learning techniques (Random Forest in particular), tu streamline the image analysis of large root dataset.
PRIMAL needs only a subset of the data to be analyse manual, instead of the full dataset. That subset is then used to train the algorithm behind PRIMAL the predict the parameters of interest, based on automatically acquired descriptors.
library(shiny) shiny::runGitHub("plantmodelling/primal", "plantmodelling")
A full description of the PRIMAL protocole can be found on Protocols.io:PRIMAL protocole . PRIMAL source code
Combining semi-automated root image analysis techniques with machine learning algorithms to accelerate large scale genetic root studies.Jonathan A. Atkinson, Guillaume Lobet, Manuel Noll, Markus Griffiths, Darren M. Wells
PRIMAL was developped as a collaboration between the Forchungszentrum Juelich, The Université de Liège, the Université catholique de Louvain and the University of Nottingham.